YeongHyeon/WideResNet_WRN-TF2

TensorFlow implementation of "Wide Residual Networks"

[TensorFlow 2] Wide Residual Networks

Related Repositories

ResNet-TF2
ResNeXt-TF2
ResNet-with-LRWarmUp-TF2
ResNet-with-SGDR-TF2

Concept

Concept of Wide ResNet [1].

Performance

Indicator	Value
Accuracy	0.99350
Precision	0.99353
Recall	0.99343
F1-Score	0.99347

Confusion Matrix
[[ 979    0    0    0    0    0    1    0    0    0]
 [   0 1130    0    1    1    0    1    2    0    0]
 [   0    1 1028    1    0    0    0    1    0    1]
 [   0    0    1 1008    0    1    0    0    0    0]
 [   0    0    1    0  978    0    1    0    0    2]
 [   1    0    0    5    0  884    1    0    0    1]
 [   1    2    0    0    1    1  953    0    0    0]
 [   0    1    5    0    0    0    0 1021    1    0]
 [   5    0    4    3    1    0    1    1  957    2]
 [   0    0    2    0    4    3    0    3    0  997]]
Class-0 | Precision: 0.99290, Recall: 0.99898, F1-Score: 0.99593
Class-1 | Precision: 0.99647, Recall: 0.99559, F1-Score: 0.99603
Class-2 | Precision: 0.98751, Recall: 0.99612, F1-Score: 0.99180
Class-3 | Precision: 0.99018, Recall: 0.99802, F1-Score: 0.99408
Class-4 | Precision: 0.99289, Recall: 0.99593, F1-Score: 0.99441
Class-5 | Precision: 0.99438, Recall: 0.99103, F1-Score: 0.99270
Class-6 | Precision: 0.99478, Recall: 0.99478, F1-Score: 0.99478
Class-7 | Precision: 0.99319, Recall: 0.99319, F1-Score: 0.99319
Class-8 | Precision: 0.99896, Recall: 0.98255, F1-Score: 0.99068
Class-9 | Precision: 0.99402, Recall: 0.98811, F1-Score: 0.99105

Total | Accuracy: 0.99350, Precision: 0.99353, Recall: 0.99343, F1-Score: 0.99347

Requirements

• Python 3.7.6
• Tensorflow 2.1.0
• Numpy 1.18.1
• Matplotlib 3.1.3

Reference

[1] Sergey Zagoruyko et al. (2016). Wide Residual Networks. arXiv preprint arXiv:1605.07146.